For example, in mobile wireless communication systems such as LTE (Long Term Evolution), LTE-Advanced, and WiMAX, each of a base station and a mobile terminal is provided with a plurality of transmission/reception antennas, and is capable of achieving high-speed data transmission by MIMO (Multi Input Multi Output) techniques. Meanwhile, by using a reference signal for channel state measurement, a mobile terminal estimates a channel state between the base station and the mobile terminal, adaptively controls the modulation scheme and coding rate (MCS (Modulation and Coding Scheme)), the number of spatial multiplexing (layers, rank), precoding weights (precoding matrix) and the like based on the estimation result, and is thereby capable of achieving more efficient data transmission. For example, it is possible to use the method as described in Non-patent Document 1.
Meanwhile, in the case of using multicarrier transmission schemes such as an OFDM (Orthogonal Frequency Division Multiplexing) scheme and OFDMA (Orthogonal Frequency Division Multiple Access) scheme as a transmission scheme, it is possible to use reference signals scattered in resource elements (each of which is an element comprised of one subcarrier in one OFDM symbol) in the frequency domain and the time domain as a reference signal for channel state measurement specific to a base station. As feedback information to estimate using such a reference signal for channel state measurement, it is possible to use information based on the channel state (explicit CSI (Channel State Information)), recommended transmission format information (implicit CSI (for example, including the CQI (Channel Quality Indicator), RI (Rank Indicator), PMI (Precoding Matrix Index), etc.) for the base station, and the like.
Particularly, since the explicit CSI is information based on the actual channel state, the feedback information amount is large as compared with the implicit CSI that is index information mainly based on a code book. Therefore, in Non-patent Documents 2 and 3 are studied techniques for reducing the information amount of explicit CSI, and for example, studied are techniques using eigenvalue decomposition, orthogonal transform such as DCT (Discrete Cosine Transform), vector quantization, and the like.